Springe direkt zu Inhalt

Guest-Talk: "Human-Aided Artificial Intelligence, or: How to Run Large Computations in Human Brains?" by Rainer Mühlhoff, May 23rd

News from May 11, 2019

May 23rd, 2 pm, Rainer Mühlhoff will give the talk: "Human-Aided Artificial Intelligence, or: How to Run Large Computations in Human Brains? - On the role of HCI for Deep Learning" at HCC Lab. The guest talk is part of the research seminar: "Human-Centered Computing".

Rainer Mühlhoff, Dr. phil., is a researcher fellow in philosophy at the Collaborative Research Center 1171 "Affective Societies" at Freie Universität Berlin. His research areas are social philosophy, critical social theory of the digital society, and philosophy of media and technology. He studied mathematics, theoretical physics, philosophy and gender studies in Heidelberg, Münster, Leipzig and Berlin. In earlier stages of his life he worked as a programmer and management consultant.

Abstract

In the past 10 years, the success of Deep Learning has led to a boom in artificial intelligence (AI) in many industrial and commercial sectors. In this talk I will argue that this upheaval has not only been driven by technological progress (better algorithms and high performance computing), but is linked to a comprehensive structural change in media culture. Current Machine Learning technologies are largely based on harnessing human cognitive resources in hybrid human-machine apparatuses. Human-machine interfaces and media infrastructures have been developed to capture human participation in order to generate a constant stream of training or verification data for various AI problems. Artificially intelligent apparatus are in fact hybrid human-machine networks. I will refer to this technological, economic and media-sociological constellation as “HumanAided AI”. Distinguishing classical AI from ‘cybernetic AI’, I will argue that AI today does not so much resemble the popular notion of intelligence as an autonomous, sovereign and rational intellectual capacity, but refers to a relational information processing capacity of hybrid human-machine networks. I will analyze this constellation with respect to questions of end-user subjectification and governmentality theory.

Date: May 23, 2019
Location: Königin-Luise-Straße 24-26, 14195 Berlin, room 120

37 / 100